Real-time packet video transmission over wireless networks isexpected to experience bursty packet losses that can causesubstantial degradation to the transmitted video quality. Inwireless networks, channel state information is hard to obtain ina reliable and timely manner due to the rapid change of wirelessenvironments. However, the source motion information is alwaysavailable and can be obtained easily and accurately from videosequences. Therefore, in this paper, we propose a novelcross-layer framework that exploits only the motion informationinherent in video sequences and efficiently combines apacketization scheme, a cross-layer forward error correction(FEC)-based unequal error protection (UEP) scheme, an intracodingrate selection scheme as well as a novel intraframe interleavingscheme. Our objective and subjective results demonstrate that theproposed approach is very effective in dealing with the burstypacket losses occurring on wireless networks without incurring anyadditional implementation complexity or delay. Thus, thesimplicity of our proposed system has important implications forthe implementation of a practical real-time video transmissionsystem.
ASJC Scopus subject areas
- Signal Processing
- Hardware and Architecture
- Electrical and Electronic Engineering